CN110446577A - Welded condition decision-making system and welded condition determination method - Google Patents

Welded condition decision-making system and welded condition determination method Download PDF

Info

Publication number
CN110446577A
CN110446577A CN201880015443.4A CN201880015443A CN110446577A CN 110446577 A CN110446577 A CN 110446577A CN 201880015443 A CN201880015443 A CN 201880015443A CN 110446577 A CN110446577 A CN 110446577A
Authority
CN
China
Prior art keywords
welded condition
welding
abnormality degree
value
physical quantity
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201880015443.4A
Other languages
Chinese (zh)
Other versions
CN110446577B (en
Inventor
藤井达也
大根努
福永敦史
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jinsyo System Co Ltd
Kobe Steel Ltd
Original Assignee
Jinsyo System Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jinsyo System Co Ltd filed Critical Jinsyo System Co Ltd
Publication of CN110446577A publication Critical patent/CN110446577A/en
Application granted granted Critical
Publication of CN110446577B publication Critical patent/CN110446577B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/095—Monitoring or automatic control of welding parameters
    • B23K9/0953—Monitoring or automatic control of welding parameters using computing means
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups
    • B23K31/003—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups relating to controlling of welding distortion
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups
    • B23K31/12—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups relating to investigating the properties, e.g. the weldability, of materials
    • B23K31/125—Weld quality monitoring
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/095—Monitoring or automatic control of welding parameters
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/095—Monitoring or automatic control of welding parameters
    • B23K9/0956—Monitoring or automatic control of welding parameters using sensing means, e.g. optical
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
    • G05B19/406—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by monitoring or safety
    • G05B19/4063—Monitoring general control system
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/34—Director, elements to supervisory
    • G05B2219/34465—Safety, control of correct operation, abnormal states
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/45—Nc applications
    • G05B2219/45104—Lasrobot, welding robot

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Plasma & Fusion (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Manufacturing & Machinery (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Human Computer Interaction (AREA)
  • Manipulator (AREA)
  • Numerical Control (AREA)
  • Arc Welding Control (AREA)
  • Butt Welding And Welding Of Specific Article (AREA)

Abstract

Welded condition decision-making system is obtained with the period corresponding to wobbling action and the relevant characteristic quantity of the physical quantity that changes in every swing period.In addition, calculating abnormality degree based on the average and standard deviation of the i.e. multiple past values of the characteristic quantity obtained about the characteristic quantity obtained in one cycle i.e. observation and before a cycle.The abnormality degree of calculating is then based on to determine welded condition.

Description

Welded condition decision-making system and welded condition determination method
Technical field
The present invention relates to the welded condition decision-making systems and welded condition determination method that determine the welded condition in arc welding.
Background technique
As the method for determining the welded condition in arc welding, there is method disclosed in patent document 1.In 1 disclosure of patent document Method in, by weldingvoltage, welding current, speed of welding, welding wire feeding speed, spray gun height and inert gas flow Each detected value is as welding signal, and in the case where one of welding signal independent change, determining to each welding signal can be good The 1st phase threshold for keeping the quality of welding layer well, in the case where 2 welding signals change simultaneously, to the complete of welding signal Portion's combination determination can well keep the range of the detected value of the quality of welding layer.Determine welding layer quality when, select 1 or The welding signal of selection is compared with the range of the 1st phase threshold or detected value, compares knot according to it by multiple welding signals Fruit determines whether defect occur in welding layer.
Existing technical literature
Patent document
Patent document 1: No. 3906561 bulletins of Japanese Patent No.
Summary of the invention
Subject to be solved by the invention
But in the method disclosed in patent document 1, the feelings in welding conditions such as change welding object, gas, setting electric currents are needed Its range for changing the 1st phase threshold and detected value is corresponded under condition.In particular, there is the detected value of the number of combinations of welding signal Range, it is necessary to by its all change.
The present invention is proposed in view of the situation, its main purpose is, setting can not be changed welding condition change by providing Value ground high-precision determines the welded condition decision-making system and welded condition determination method of welded condition.
Means for solving the problems
In order to solve above-mentioned problem, the welded condition decision-making system of a scheme of the invention is to making welding gun periodical wobble The welded condition in arc welding under wobbling action is determined have: acquisition unit, in each week of the wobbling action Phase obtains characteristic quantity relevant to the physical quantity changed corresponding to the period；Calculation section, based on about by the acquirement Characteristic quantity, that is, the observation and obtained before one period by the acquisition unit that portion was obtained period The characteristic quantity, that is, multiple past values are average and standard deviation calculates abnormality degree；And determination unit, based on by described The abnormality degree that calculation section calculates determines the welded condition.
It, can also be with the calculation section is configured to calculate the abnormality degree, the welding in every period on the basis of the program State decision-making system is also equipped with: except outside, the abnormality degree calculated in the period before one period is than given A reference value it is high in the case where, will be described before period in the abnormality degree calculating in be used as described in the observation Characteristic quantity is except the multiple past value.
In addition, on the basis of above scheme, it can also be with the welded condition decision-making system is also equipped with: approximate portion, by the spy The time fluctuation of sign amount is approximate, and the calculation section is configured to based on the observation and the observation that is obtained by the approximate portion The difference of the approximation of the difference of the approximation of value and the past value and the past value obtained by the approximate portion it is average with And standard deviation calculates the abnormality degree.
, can also be in addition, on the basis of above scheme, the approximation portion is configured to the characteristic quantity through linear approximation Time fluctuation is approximate.
, can also be in addition, on the basis of above scheme, the approximation portion is configured to by the way that 2 ranks are approximate or exponential approximation is by institute The time fluctuation for stating characteristic quantity is approximate.
, can also be in addition, on the basis of above scheme, the characteristic quantity be average value in the period of the physical quantity, The physical quantity of the welding gun in the wobbling action waved at end, the welding gun in the wobbling action wave The physical quantity of centre, comprising wave end section in the physical quantity integrated value or week with the physical quantity The amplitude of the approximate waveform of phase property time fluctuation.
In addition, on the basis of above scheme, it can also be with the acquisition unit is constituted are as follows: is obtained in the period of the physical quantity The average value, physical quantity waved at end of the welding gun in the wobbling action, described in the wobbling action The physical quantity for waving centre of welding gun, comprising wave end section in the physical quantity integrated value or with it is described In the amplitude of the approximate waveform of periodical time fluctuation of physical quantity more than at least two, as the characteristic quantity, the calculating Portion is configured to calculate each characteristic quantity respectively the abnormality degree, and the determination unit is configured to be based on being calculated by the calculation section Each abnormality degree out determines the welded condition.
In addition, on the basis of above scheme, it can also be with the welded condition decision-making system is also equipped with: control unit, based on described The judgement result of the welded condition of determination unit controls the movement of the arc welding.
, can also be with the welded condition decision-making system is also equipped in addition, on the basis of above scheme: exceptional communication portion, by In the case that the determination unit is determined as that the welded condition is exception, the exception is notified.
In addition, on the basis of above scheme, it can also be with the physical quantity is welding current or weldingvoltage.
In addition, the welded condition determination method of other schemes of the invention is to the electricity under the wobbling action for making welding gun periodical wobble Welded condition in arc welding is determined, with following steps: being obtained and is corresponded in each period of the wobbling action The period and the relevant characteristic quantity of physical quantity changed；Based on being seen about the characteristic quantity obtained in a period The average and standard deviations of measured value and the i.e. multiple past values of the characteristic quantity obtained before one period calculates Abnormality degree；With the welded condition is determined based on the abnormality degree of calculating.
The effect of invention
Related welded condition decision-making system and welded condition determination method according to the present invention, even if welding condition changes, Also without change setting high-precision it can determine welded condition.
Detailed description of the invention
Fig. 1 is the schematic diagram for indicating the structure of welding system involved in embodiment 1.
Fig. 2 is the block diagram for indicating the structure of robot controller.
Fig. 3 is the block diagram for indicating the structure of welded condition decision-making system involved in embodiment 1.
Fig. 4 is the flow chart for indicating the running order of robot controller and power supply device.
Fig. 5 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 1.
Fig. 6 A is the explanatory diagram for illustrating the variation of the welding wire length in wobbling action.
Fig. 6 B is the explanatory diagram for illustrating the variation of the welding wire length in wobbling action.
Fig. 7 is the perspective view for indicating workpiece used in evaluation test 1.
Fig. 8 A is the chart for indicating the average current measured in evaluation test 1.
Fig. 8 B is the chart for indicating the abnormality degree calculated in evaluation test 1.
Fig. 9 is the perspective view for indicating workpiece used in evaluation test 2.
Figure 10 A is the chart for indicating the average current measured in evaluation test 2.
Figure 10 B is the chart for indicating the abnormality degree calculated in evaluation test 2.
Figure 11 A is the chart for indicating the average current measured in evaluation test 3.
Figure 11 B is the chart for indicating the abnormality degree calculated in evaluation test 3.
Figure 12 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 2.
Figure 13 is the chart for indicating the abnormality degree calculated in evaluation test 4.
Figure 14 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 3.
Figure 15 A is the chart for indicating the left end electric current measured in evaluation test 5.
Figure 15 B is the chart for indicating the abnormality degree about the left end electric current calculated in evaluation test 5.
Figure 16 A is the chart for indicating the right end electric current measured in evaluation test 5.
Figure 16 B is the chart for indicating the abnormality degree about the right end electric current calculated in evaluation test 5.
Figure 17 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 4.
Figure 18 A is the chart for indicating the amplitude calculated in evaluation test 6.
Figure 18 B is the chart for indicating the abnormality degree about the amplitude calculated in evaluation test 6.
Figure 19 A is the chart for indicating the amplitude calculated in evaluation test 7.
Figure 19 B is the chart for indicating the abnormality degree about the amplitude calculated in evaluation test 7.
Figure 20 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 5.
Figure 21 is the histogram of the average current in evaluation test 2.
Figure 22 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 6.
Figure 23 is the chart for indicating the result of the linear approximation in evaluation test 8.
Figure 24 A is the histogram of the difference of the average current and approximation in evaluation test 8.
Figure 24 B is the chart for indicating the abnormality degree calculated in evaluation test 8.
Figure 25 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 7.
Figure 26 is the chart for indicating the abnormality degree calculated in evaluation test 9.
Figure 27 is the flow chart for indicating the running order of welded condition decision-making system involved in embodiment 8.
Figure 28 is the chart for indicating the time fluctuation of the average current after the rigid starting the arc.
Figure 29 A is the chart for indicating the approximate result of 2 ranks in evaluation test 10.
Figure 29 B is the chart for indicating the abnormality degree calculated in evaluation test 10.
Figure 30 is the explanatory diagram for illustrating the welding in the copying control to the workpiece for implementing tack welding.
Figure 31 A is the chart for illustrating an example of the integrated value comprising waving the welding current in the fixed interval at end.
Figure 31 B is other exemplary charts for illustrating the integrated value comprising waving the welding current in the fixed interval at end.
Specific embodiment
Illustrate the preferred embodiments of the present invention below with reference to attached drawing.In addition, each embodiment as shown below instantiates For the method for technical idea materialization of the invention and device, technical idea of the invention to be not limited to following.This The technical idea of invention can add various changes in the technical scope that claims are recorded.In addition, as shown below each It enumerates in embodiment and is illustrated for the welding robot of manipulator, but it's not limited to that for applicable object of the invention, It can also be using the automatic soldering device other than manipulator as applicable object.
(embodiment 1)
In the present embodiment, welded condition decision-making system obtains welding current value to each swing period in each period of swing Average value as characteristic quantity, abnormality degree is calculated to this feature amount, welded condition is determined based on abnormality degree.
<structure of welding system>
Fig. 1 is the schematic diagram for indicating the structure of welding system involved in present embodiment.Welding system 10 has welding robot People 20, robot controller 30, power supply device 40 and welded condition decision-making system 100.Welded condition decision-making system 100 determines The welded condition of welding robot 20.
Welding robot 20 is made of the manipulator of vertical joint type, has welding gun 21 in its front end.Involved by present embodiment And welding robot 20, carry out MIG (Metal Inert Gas, welding) welding or MAG (Metal Active Gas, metal active gas welding) welding etc. consumable electrodes formula arc welding.Welding robot 20 and robot Control device 30 and power supply device 40 are separately connected.
In welding gun 21, it is admitted to welding wire 24 from wire feed unit 23, is sent out from the front end of welding gun 21 by welding wire 24.Power supply device 40 have constant voltage power supply 41, welding current test section 42 and switch 43.After switch 43 is connected, to welding wire 24 from constant voltage power supply 41 provide electric power.Thus to weldingvoltage is applied between welding wire 24 and workpiece (welded part) 50, electric arc is generated.Switch 43 disconnects Afterwards, stop providing to the electric power of welding wire 24, stop the generation of electric arc.In addition, the detection of welding current test section 42 produces in welding Raw welding current.
Power supply device 40 also has CPU44, memory 45 and communication unit 46.The computer program of power supply control, that is, power supply control Program 47 is stored in memory 45, executes the power supply control program 47 by CPU44, to carry out the on/off etc. of switch 43 The control of weld power.In addition, power supply device 40 is connect with wire feed unit 23, CPU44 by power supply control program 47 come Control the feed speed of welding wire.Communication unit 46 is the communication module of wired or wireless communication.The communication unit 46 uses given Communication protocol carries out data communication between robot controller 30 and welded condition decision-making system 100.
The structure of robot controller 30 will be illustrated next.The movement of the control welding robot 20 of robot controller 30. Fig. 2 is the block diagram for indicating the structure of robot controller 30.Robot controller 30 has CPU301, memory 302, packet Operation panel 303, teaching machine 304, input and output portion 305 and communication unit 306 containing multiple switch.
The control program 330 of the computer program of control as welding robot 20, is stored in memory 302, passes through CPU301 executes the control program 330, the control of the welding movement of Lai Jinhang welding robot 20.
Operation panel 303 and teaching machine 304 are used in the input of the instruction to robot controller 30.Operator's energy Tutorial program is inputted to teaching machine 304.Robot controller 30 follows the tutorial program inputted from teaching machine 304 and welds to control Welding robot 20.Tutorial program can also not shown computer be made.In this case, movable-type recording medium can be passed through Handover, or by data communication, tutorial program is supplied to robot controller 30.
(do not scheme in the driving circuit of 305 connecting welding current detecting part 42 of input and output portion and the actuator of welding robot 20 Show).The value of the welding current detected by welding current test section 42 is supplied to CPU301 via input and output portion 305. CPU301 carries out the control of aftermentioned such welding robot 20 by controlling program 330, and control signal is output to bonding machine The driving circuit of device people 20.
Communication unit 306 is the communication module of wired or wireless communication.The communication unit 306 using given communication protocol come with Data communication is carried out between power supply device 40 and welded condition decision-making system 100.
The robot controller 30 of above structure, control welding robot 20 each axis come control welding gun 21 position and From the protrusion length (hereinafter referred to as " welding wire length ") of the welding wire 24 of welding gun 21.In welding movement, correspond to welding gun 21 and work The distance between part 50 adjusts welding wire length.That is, if welding gun 21 is reduced by welding wire length close to workpiece 50, if welding gun 21 from Workpiece 50 deviates from, and just increases welding wire length.Corresponding to welding wire length length and there are the resistance variations in welding wire 24, the resistance Variation causes the variation of welding current value.For this purpose, robot controller 30 by the feedback control using welding current value come Position and the welding wire length for adjusting welding gun 21, so that welding current value becomes suitable.
In addition, robot controller 30 makes welding robot 20 execute wobbling action.Wobbling action is handed over welding direction Welding direction (is referred to as " front " below, right when observing the front is referred to as " right ", when by observation front by the direction of fork Left be referred to as " left ") on make the alternately movement waved of welding gun 21.Robot controller 30 controls welding robot 20, makes It obtains it and wobbling action is carried out by swing period, amplitude, speed of welding of setting.
In turn, robot controller 30 and above-mentioned wobbling action execute sealing wire copying control together.Weld wire modeling control System is to control the left-right position of welding gun 21 along the movement of sealing wire formation welding bead.
The structure of welded condition decision-making system 100 will be illustrated next.Fig. 3 is to indicate that welded condition involved in present embodiment is sentenced Determine the block diagram of the structure of system 100.Welded condition decision-making system 100 is realized by computer 101.Computer 101 has main body 110, input unit 120 and display unit 130.Main body 110 has CPU111, ROM112, RAM113, hard disk 114, input/output interface 115, communication interface 116 and image output interface 117.CPU111, ROM112, RAM113, hard disk 114, input/output interface 115, communication interface 116 and image output interface 117 are connected with each other by bus.
The welded condition decision procedure as the computer program for determining the state of arc welding is executed by CPU111 150, computer 101 is functioned as welded condition decision-making system 100.
It is had recorded in ROM112 in the CPU111 computer program executed and its used data etc..RAM113 is for reading It is recorded in the welded condition decision procedure 150 of hard disk 114.RAM113 is when CPU111 executes computer program, as CPU111 Operating area and utilize.
Hard disk 114 be mounted with operating system and application program etc. for make CPU111 execute various computer programs and Data used in the execution of the computer program.Welded condition decision procedure 150 is also mounted at the hard disk 114.
The input unit 120 being made of keyboard and mouse is connected in input/output interface 115.It is connected in input/output interface 115 Welding current test section 42.The value of the welding current detected by welding current test section 42 is mentioned via input/output interface 115 Supply CPU111.CPU111 determines the state of arc welding using welding current value by welded condition decision procedure 150.
Communication interface 116 is the communication module of wired or wireless communication.The communication interface 116 is come using given communication protocol Data communication is carried out between power supply device 40 and robot controller 30.Image output interface 117 with by LCD or CRT Etc. compositions display unit 130 connect, video signal corresponding with the image data provided from CPU111 is output to display unit 130.Display unit 130 follows the video signal of input to show image (picture).
<movement of welding system>
The movement of welding system 10 will be illustrated next.In the case where carrying out arc welding, robot is respectively started in operator Control device 30, power supply device 40 and welded condition decision-making system 100.Robot controller 30 controls welding robot 20 Movement, power supply device 40 to welding robot 20 provide weld power.In addition, 100 real-time judgment of welded condition decision-making system The state of the arc welding of welding robot 20.
The movement of robot controller 30 and power supply device 40 is illustrated.Fig. 4 is to indicate robot controller 30 And the flow chart of the running order of power supply device 40.Operator operates teaching machine 304 in the case where starting arc welding To input robot controller 30 tutorial program, various setting values, instruction welding starts.Robot controller 30 CPU301 receives the tutorial program, various setting values and welding and starts to indicate (step S101).
CPU301 instructs the offer of weld power to start (step S102) power supply device 40.Power supply device 40 receives the instruction Switch 43 is connected (step S104) by (step S103), CPU44.Thus apply weldingvoltage between welding wire 24 and workpiece 50, Generate electric arc.
Next, CPU301 butt welding machine device people 20 sends control signal, welding control (step S105) is executed.It is controlled in welding In comprising wobbling action control (step S1051) and sealing wire copying control (step S1052).In addition, being controlled in welding In, also carry out the position using the welding gun 21 of welding current value and the feedback control of welding wire length.
CPU301 determines a need for the stopping (step S106) of welding movement.In the case where not needing the stopping of welding movement ("No" in step S106), the process returns to step S105 by CPU301, gradually execute welding control.
Such as the receiving in the instruction for thering is the welding from operator to stop, the welding of the sensor set on welding robot 20 The detection of end position, the detection of abnormal welding and the welding from welded condition decision-making system 100 stop the receiving of request Any one in the case where, CPU301 be judged as need welding act stopping ("Yes" in step S106), will welding control stop Only (step S107) instructs the offer of weld power to stop (step S108) power supply device 40.Power supply device 40 receives this and refers to It enables (step S109), switch 43 is disconnected (step S110) by CPU44, is mentioned to stop the weld power of butt welding machine device people 20 For.Thus the movement of robot controller 30 and power supply device 40 terminates.
The movement of welded condition decision-making system 100 will be illustrated next.Fig. 5 is the movement time for indicating welded condition decision-making system 100 The flow chart of sequence.Firstly, variable i of the CPU111 of welded condition decision-making system 100 to the period (number) for indicating wobbling action 0 (step S111) of set initial value.
In power supply device 40, welding current test section 42 persistently detects welding current value, exports as detected value. CPU111 receives the time series data (step S112) of welding current value from welding current test section 42.
Illustrate welding current value herein.In wobbling action, the left and right variation relative to welding direction of the position of welding wire 24.At this In the wobbling action of sample, the change in location of the left and right directions corresponding to welding wire 24, the welding position of the front end of welding gun 21 and workpiece 50 The distance set changes, and changes corresponding to this welding wire length.At this moment, welding wire length is controlled, so that the front end of welding wire 24 With at a distance from the welding position of workpiece 50 i.e. arc length be it is approximately fixed.Fig. 6 A and Fig. 6 B are for illustrating in wobbling action Welding wire length variation explanatory diagram.In the case where workpiece 50 is banjo fixing butt jointing, as shown in figure 6 a the two of docking Groove 51 is set between workpiece 50.For this purpose, when the workpiece 50 when welding wire 24 is opposed with groove 51 and with docking is opposed, welding gun 21 change at a distance from workpiece 50.It changes for this purpose, corresponding to this apart from welding wire length.In addition, workpiece 50 be T connector, In the case where cruciform joint, angle joint etc., as shown in fig. 6b, two workpiece 50 are configured to intersect.The cross part becomes Sealing wire.When welding wire 24 is opposed with cross part (interconnecting pieces of two workpiece 50), welding wire length longest, as welding wire 24 is from the friendship Fork mobile welding wire length on the direction (arrow direction in figure) intersected with welding direction shortens.Welding current value is corresponding thus In wobbling action period and change.
Fig. 5 is referred to once again.CPU111 calculates average value (hereinafter referred to as " average current ") conduct of the welding current value in the i period Characteristic quantity (step S113).CPU111 determines whether the data number i of the characteristic quantity calculated reaches given number N (step S114), In the case where data number i is less than N ("No" in step S114), i is set to be incremented by (step S115), the process returns to step S112. Thus characteristic quantity is respectively obtained in the continuous multicycle.
On the other hand, in the case where data number i is N or more ("Yes" in step S114), CPU111 calculates abnormality degree (step S116).Abnormality degree is defined by following formula.
[mathematical expression 1]
Here, so-called observation, refers to the characteristic quantity in concern period (being in this case i, the i.e. newest period).In addition, flat Mean value is the average value of the multiple characteristic quantities (hereinafter referred to as " past value ") obtained before paying close attention to the period, and standard deviation is the past The standard deviation of value.
The abnormality degree in present embodiment is further described.About the average current Iave in the i-th period of wobbling action (i), the average value of past value is provided in formula below (1), and standard deviation is provided in formula (2), and abnormality degree is provided in formula (3).
[mathematical expression 2]
That is, average value is that closely the past value of the data number N before period i is averaged, standard deviation is the past value of data number N Standard deviation.Such abnormality degree is to indicate to show the statistical of different degree relative to N number of past value observation Numerical value.Here, welding current value, since the variation in each period is big, if N is too small, average and standard deviation value will Become unstable, can no longer calculate the abnormality degree of correct reflection welded condition.N is preferably larger thus.On the other hand, if due to N It is excessive, will be elongated during being used to obtain data, therefore practical preferred N is smaller.Accordingly, and specifically, it is preferable to N be 30 with Upper 100 or less.Especially using Ar-CO2In the case where the few protective gas of the such noise of mixed gas, preferably N is set as 30 or more 50 or less.If but have the problem of precision, consider N being set as 50 or more.
Following CPU111 determines welded condition (step S117) based on the abnormality degree of calculating.In this process, CPU111 root The stopping of welding movement is determined a need for according to abnormality degree.If abnormality degree is more than given a reference value, welded condition is different Often, it is judged to needing to weld the stopping of movement.On the basis of abnormality degree value hereinafter, if welded condition be it is normal, be judged to being not required to Weld the stopping of movement.The a reference value can be provided by operator, can also be set automatically with CPU111.Automatically it is setting In, it can be using the given value in chi square distribution as a reference value.The given value can consider the abnormal incidence in welding to determine. If such as the abnormal probability of happening in welding is 0.5%, 0.5% in chi square distribution can be set as a reference value.
CPU111 (" normal " in step S117) in the case where being determined as that welded condition is normal situation, is passed to step for processing S115.On the other hand, CPU111 ("abnormal" in step S117) in the case where being determined as that welded condition is exception, makes display unit The picture (step S118) that 130 display notices occur extremely sends welding to robot controller 30 and stops request (step S119).Welding provisioning request is received by robot controller 30, welding control stops, the welding electricity of butt welding machine device people 20 The offer of power stops.Terminate the movement of welded condition decision-making system above.
By constituting as described above, the abnormality degree that welded condition decision-making system 100 calculates welded condition is used as relative to N number of mistake Remove the statistical numerical value of the observation of value.Such abnormality degree is being welded due to only being determined with past value and observation It is unaffected in the welding conditions such as object, protective gas, setting electric current.Therefore, even if welding condition changes, welded condition determines System 100 also can not the ground such as change setting calculate abnormality degree corresponding with the welding condition at the time point, can accurately sentence Disconnected welded condition.
As described above, computer 101 is functioned as following each portion: being obtained in each period and is corresponded to wobbling action The acquisition unit of the relevant characteristic quantity of the physical quantity of cyclical swing；The calculation of abnormality degree is calculated according to the characteristic quantity obtained by the acquisition unit Portion out；The determination unit of welded condition is determined based on the abnormality degree calculated by the calculation section；It is being determined as that welded condition is abnormal In the case where notify abnormal abnormal notification unit.In addition, robot controller 30 is as the welded condition based on determination unit Determine that result functions to control the control unit of the movement of arc welding.
<evaluation test 1>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.Fig. 7 is the perspective view for indicating workpiece 50 used in this evaluation test.In this evaluation test, make as workpiece 50 With the T connector being made of vertical plate 52 and lower plate 53.Welding is set as the fillet welded to the cross part of vertical plate 52 and lower plate 53 Welding.Vertical plate 52 is compared with lower plate 53 and the length of welding direction (longitudinal direction) is shorter, in the welding direction of lower plate 53 It is short of on way.In this evaluation test, welded from there are the parts of vertical plate 52 to no longer existing part, to verify whether energy It is the variation of welded condition by the end point detection of vertical plate 52.In this evaluation test, hunting frequency is set as 2Hz, by data number N It is set as 40.
The average current Iave in each period of wobbling action is calculated, calculates average current using above-mentioned formula (1)~(3) Abnormality degree.Fig. 8 A is the chart for indicating the average current Iave measured in this evaluation test.In fig. 8 a, horizontal axis indicates to swing dynamic The number (being set as 1 period 1 time) of work, the longitudinal axis indicate average current Iave.As shown in Figure 8 A like that, in number of oscillations 1~57, Average current is about 270~275A, and in contrast, after number of oscillations the 59th time, average current is reduced to 267A or less.It should 59th time corresponding with the end of vertical plate 52, and the result of Fig. 8 A is indicated since there is no vertical plate 52, average currents after the 59th time It reduces.
Fig. 8 B is the chart for indicating the abnormality degree calculated in this evaluation test.In the fig. 8b, horizontal axis indicates the number of wobbling action, The longitudinal axis indicates abnormality degree.As shown in Figure 8 B like that, at the 1st~57th time, abnormality degree is about 0, in contrast, the 59th~the 63 times, abnormality degree is more than a reference value.In this evaluation test, 0.5% value that a reference value is set as in chi square distribution 7.879439.It, can since the number of oscillations that the abnormality degree is more than a reference value is the 59th time corresponding with the terminal of vertical plate 52 Know the variation that the end of vertical plate 52 can be correctly detected as welded condition.In addition, abnormality degree is lower than once again after the 64th time A reference value.It is thought that because the high data of abnormality degree for welding the part of no vertical plate 52 are included in past value.
<evaluation test 2>
Next, inventors carry out implementation evaluation test using the T connector for implementing tack welding as workpiece.Fig. 9 is to indicate originally to comment The perspective view of workpiece 50 used in valence test.In this evaluation test, as workpiece 50, using to by vertical plate 52 and lower plate 53 The cross part of the T connector of composition implemented the workpiece of the tack welding 54 of output.Welding is set as intersecting to vertical plate 52 and lower plate 53 The fillet weld that portion is welded.In this evaluation test, along the part of cross part never tack welding 54 in the presence of positioning The part of weldering 54 is welded, to verify whether that tack welding 54 can be detected as to the variation of welded condition.In addition, being tried in this evaluation In testing, hunting frequency is set as 2Hz identically as evaluation test 1, data number N is set as 50.In addition, with same with evaluation test 1 The gimmick of sample is tested.
Figure 10 A is the chart for indicating the average current Iave measured in this evaluation test.In Figure 10 A, horizontal axis indicates to swing dynamic The number of work, the longitudinal axis indicate average current Iave.As shown in Figure 10 A like that, in number of oscillations 1~72, average current is about 270 ~275A, in contrast, after number of oscillations the 73rd time, average current steeply increases.73rd time with tack welding 54 Position is corresponding, and the result of Figure 10 A is indicated due to the average current increase there are tack welding 54 after the 73rd time.That is, in the number In, influenced and average current large change after the 73rd time by tack welding 54.
Figure 10 B is the chart for indicating the abnormality degree calculated in this evaluation test.In fig. 1 ob, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate abnormality degree.As shown in Figure 10 B like that, at the 1st~72nd time, on the basis of abnormality degree value hereinafter, in contrast, At the 73rd~77th time, abnormality degree is more than a reference value.Since the number of oscillations that the abnormality degree is more than a reference value is and tack welding 54 Position it is 73rd time corresponding, thus, it can be known that tack welding 54 can be correctly detected to the variation for welded condition.
<evaluation test 3>
Next, inventors burn workpiece by the way that target current value is set to height, verify whether the burn-through can be examined Survey the variation for welded condition.In addition, hunting frequency is set as 2Hz identically as evaluation test 1 in this evaluation test, it will Data number N is set as 10.In addition, being tested with gimmick same as evaluation test 1.
Figure 11 A is the chart for indicating the average current Iave measured in this evaluation test.In Figure 11 A, horizontal axis indicates to swing dynamic The number of work, the longitudinal axis indicate average current Iave.As shown in Figure 11 A like that, in number of oscillations 1~18, average current is with about 350 ~320A passage.In this evaluation test, burnt at number of oscillations the 18th~19th time.Although being influenced by the burn-through And reduced in the 19th average current, but its reduction amount cannot be said greatly, it is difficult to be burnt with average current detection.
Figure 11 B is the chart for indicating the abnormality degree calculated in this evaluation test.In Figure 11 B, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate abnormality degree.As shown in Figure 11 B like that, at the 1st~17th time, on the basis of abnormality degree value hereinafter, in contrast, At the 18th~19th time, abnormality degree is more than a reference value.It is found that being hair since the abnormality degree occurs more than the number of oscillations of a reference value Wear the 18th and the 19th time of raw burn, therefore can will burn the variation that correctly detect as welded condition.
(embodiment 2)
In the welded condition determination method involved in embodiment 1, although if welded condition change dramatically, abnormality degree will one Degree is got higher, but in the case where the welded condition of the exception continues, the high data of abnormality degree will be included in past value, sometimes The abnormality degree calculated using the past value can reduce (with reference to Fig. 8 B, Figure 10 B, Figure 11 B).In the present embodiment, welded condition Decision-making system by past abnormality degree be more than given a reference value characteristic quantity from past value except external calculating is average and standard deviation Difference calculates abnormality degree using the average and standard deviation.As a result, in the case where abnormal welded condition continues, it can connect Continuous detection is abnormal.
<structure of welding system>
The structure of the welding system as involved in present embodiment and the structure of welding system 10 involved in embodiment 1 are same Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, same due to illustrating with embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 12 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~115 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
If average current when being more than a reference value comprising the abnormality degree calculated in N number of past value, in the past, CPU111 are just removed (step S2161) outside.In the case where abnormality degree is more than a reference value, it is abnormal for capable of being estimated as welded condition.Therefore, by by this The high data of the abnormality degree of sample remove, and the average current at welded condition normal time point is only contained in past value.
Next, CPU111 selects the average value and standard deviation for eliminating the past value of the high data of abnormality degree, it is based on it Calculate abnormality degree (step S2162).It is set as in the i-th-N to (i-1)-th average current Iave that there are M abnormality degree is super Cross the average current of a reference value.The data of this M average current are set to I ' ave (i, 1), I ' ave (i, 2), I ' ave (i, 3) ..., I ' ave (i, M), they are summarized and is set as I ' ave (i).About the average current in the i-th period of wobbling action I ' ave (i) provides the average value of the past value except the high data of abnormality degree in following formula (4), and standard deviation is in formula (5) it provides, abnormality degree is provided in formula (6).
[mathematical expression 3]
CPU111 determines welded condition (step S117) based on the abnormality degree of calculating.The place that the processing and embodiment 1 illustrate Reason is same.CPU111 (is " normal ") that processing is passed to step in step S117 in the case where being determined as that welded condition is normal situation Rapid S115.On the other hand, CPU111 (being "abnormal" in step S117) in the case where being determined as that welded condition is exception, makes to show Show the picture (step S118) that the display notice of portion 130 occurs extremely.In the present embodiment, after the notice occurred extremely, CPU111 does not send welding and stops request, and processing is passed to step S115.It is being determined as that welded condition is abnormal feelings as a result, Under condition, the calculating of abnormality degree and the judgement of welded condition are also continued with.
By constituting as described above, characteristic quantity when due to being exception by welded condition, can not be by except past value The abnormality degree for influencing ground and calculating correct reflection welded condition of past abnormal characteristic quantity, it is lasting in abnormal welded condition In the case of, it can continuously detect the exception.
So, the computer 101 of this structure functions, so that calculation section above-mentioned calculates abnormality degree in each period.Separately Outside, computer 101 is also functioned as except outside, and the abnormality degree calculated in the period before a cycle is higher than given A reference value in the case where, by the abnormality degree in the period before calculating be used as observation characteristic quantity removed from multiple past values Outside.This is also the same in aftermentioned embodiment 7.
<evaluation test 4>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.Workpiece, hunting frequency and data number N and evaluation test 1 are same.
Average current when by abnormality degree more than a reference value calculates average value and standard deviation, according to it except past value Calculate abnormality degree.The data of average current used in the calculating of abnormality degree are identical as evaluation test 1.Figure 13 is to indicate this The chart of the abnormality degree calculated in evaluation test.In Figure 13, horizontal axis indicates that the number of wobbling action, the longitudinal axis indicate abnormality degree. As shown in Figure 13, after number of oscillations the 59th time, abnormality degree is more than a reference value.For the abnormality degree in evaluation test 1 It is less than after the 64th time of a reference value, abnormality degree is also above a reference value.Without vertical plate 52 after the 59th time, abnormality degree is just Really reflect the variation of the welded condition.In this way, knowing correct detection to weld shape with being influenced by the high data of abnormality degree The variation of state.
(embodiment 3)
In the present embodiment, welded condition decision-making system obtains the welding of wobbling action waved in end in each swing period Current value calculates abnormality degree to this feature amount, determines welded condition based on abnormality degree as characteristic quantity.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, same due to illustrating with embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 14 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~112 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
CPU111 is by the left end waved in the i period of wobbling action and the welding current value on right end (below by left end Welding current value is referred to as " left end electric current ", and the welding current value of right end is referred to as " right end electric current ") from the measured value of welding current It extracts, as characteristic quantity (step S313).The processing of step S114~S115 due to illustrate in embodiment 1 it is same, because The description thereof will be omitted for this.
CPU111 calculates abnormality degree (step S316) to left end electric current and right end electric current respectively.I-th week about wobbling action Each electric current of interim left end electric current IL (i) and right end electric current IR (i), the average value of past value in following formula (7) and (8) it provides, standard deviation is provided in formula (9) and (10), and abnormality degree is that formula (11) and (12) provide.
[mathematical expression 4]
Next, abnormality degree of the CPU111 based on left end electric current and right end electric current respectively about calculating, dynamic to determine to swing The left end of work and right end respectively in welded condition (step S317).In this process, CPU111 will be about left end electric current Abnormality degree and abnormality degree about right end electric current are compared with a reference value respectively, if which abnormality degree is more than a reference value, weldering The state of connecing is exactly exception, is judged to needing to weld the stopping of movement.On the basis of the abnormality degree of two sides value hereinafter, if welded condition It is normally, to be judged to not needing the stopping of welding movement.
The processing of step S118~S119 due to illustrate same in embodiment 1, the description thereof will be omitted.
, there are the situation that welded condition is different in left end and right end in the waving of wobbling action.Such as only in a part of workpiece Existing defects are that the left end in the waving of wobbling action is not passed through by the defect but right end, in this case, only left The welded condition at end becomes abnormal.For this purpose, by calculating each electric current about left end electric current and right end electric current as described above Abnormality degree, can correctly detect the variation of the welded condition in the left end of wobbling action.
<evaluation test 5>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.Workpiece, hunting frequency and data number N and evaluation test 1 are same.
The left end electric current IL and right end electric current IR in each period of wobbling action are measured, is calculated using above-mentioned formula (4)~(9) Out about left end electric current and the abnormality degree of each electric current of right end electric current.Figure 15 A is the left end for indicating to measure in this evaluation test The chart of electric current IL, Figure 16 A are the charts for indicating right end electric current IR.In Figure 15 A and Figure 16 A, horizontal axis indicates wobbling action Number, the longitudinal axis indicate welding current value.As shown in fig. 15, in number of oscillations 1~57, left end electric current is about 270~ 280A, in contrast, after number of oscillations the 58th time, left end electric current is reduced to 263A or less.Vertical plate 52 is relative to welding side To left side is located at, the welding position in the left end of wobbling action becomes on vertical plate 52.Number of oscillations the 58th time end with plate 52 Corresponding, the result of Figure 15 A indicates after the 58th time due to the left end electric current reduction without vertical plate 52.
On the other hand, about right end electric current, as shown in Figure 16 A like that, the high electricity to 281~291A is maintained under full number of oscillations Flow valuve.This indicates that the welding position in the right end of wobbling action is in lower plate 53, even if without vertical plate 52, also to lower plate 53 In welded condition do not influence.
Figure 15 B is the chart for indicating the abnormality degree about the left end electric current calculated in this evaluation test, and Figure 16 B is indicated about the right side Hold the chart of the abnormality degree of electric current.In Figure 15 B and Figure 16 B, horizontal axis indicates that the number of wobbling action, the longitudinal axis indicate abnormal Degree.As shown in fig. 15b, at the 1st~56th time, value is different at the 58~63rd time hereinafter, in contrast on the basis of abnormality degree Normal manner is more than a reference value.It is found that since the number of oscillations that the abnormality degree about left end electric current is more than a reference value is and vertical plate 52 Terminal is 58th time corresponding, therefore can correctly detect variation of the end as welded condition of vertical plate 52.On the other hand, about the right side As shown in fig 16b, continue full number of oscillations is a reference value or less to the abnormality degree of end electric current.This shows: about wobbling action Right end correctly reflect welded condition stablize passage.In addition, the abnormality degree in about left end electric current, abnormal after the 64th time Degree is lower than a reference value once again, it is believed that this is because the high data of the abnormality degree for welding the part of no vertical plate 52 are included in past value In.
(embodiment 4)
In the present embodiment, welded condition decision-making system obtains in every swing period and becomes with the periodical time of welding current value The amplitude of approximate waveform is moved as characteristic quantity, abnormality degree is calculated to this feature amount, welded condition is determined based on abnormality degree.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, same due to illustrating with embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 17 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~112 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
Time series data of the CPU111 based on the welding current value in the i period of wobbling action, come make with swing period phase It is adapted to the duplicate function of cycle with the waveform of welding current value, the amplitude of the function calculated, as feature It measures (step S413).The processing is described in detail.If amplitude is set as p1, phase is set asThe then weldering of cyclical movement The function y (t) for connecing current value energy following formula (13) carries out approximation.It is the time in this t.
[mathematical expression 5]
Y (t)=p1×sin(t+φ)+p2(13) above-mentioned formula (13) is used to the wobbling action of i period corresponding amount, such as The n formula comprising coefficient p1 and p2 is obtained down.Here, ti is the time in period i.
[mathematical expression 6]
If above-mentioned formula (14) is set as B=A × P, p1, p2 are calculated by following formula (15).
[mathematical expression 7]
P=(AT×A)-1×AT×B (15)
The obtained amplitude p1 reflects the maximum value of the welding current of 1 period of wobbling action corresponding amount.In addition, in this implementation In mode, it is set as in formula (13) using by the deviation of the phase with welding current valueSine wave function, but can also be with Use the function for sine wave and cosine wave being overlapped shown in following formula (16).
[mathematical expression 8]
Y (t)=qs×sin(t)+qc×cos(t)+p2 (16)
The amplitude qs of sine wave in formula (16) and the amplitude qc of cosine wave are sought by least squares method respectively.It uses later The synthesis of trigonometric function seeks net amplitude p1 by following formula (17).
[mathematical expression 9]
The processing of step S114~S115 due to illustrate same in embodiment 1, the description thereof will be omitted.
CPU111 calculates abnormality degree (step S416) to amplitude p1.About the amplitude p1 in the i-th period of wobbling action, past value Average value provided in following formula (18), standard deviation is provided in formula (19), and abnormality degree is provided in formula (20).
[mathematical expression 10]
Next, CPU111 determines welded condition (step S417) based on the abnormality degree of the amplitude p1 about calculating.At this In reason, the abnormality degree about amplitude p1 is compared with a reference value, if abnormality degree is more than a reference value, welded condition is exactly different Often, it is judged to needing to weld the stopping of movement.If on the basis of abnormality degree value hereinafter, welded condition be exactly it is normal, be judged to being not required to Weld the stopping of movement.
The processing of step S118~S119 due to illustrate same in embodiment 1, the description thereof will be omitted.
<evaluation test 6>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.Workpiece, hunting frequency and data number N and evaluation test 1 are same.
Using the amplitude p1 in each period of above-mentioned formula (14) and (15) to calculate wobbling action, formula (18)~(20) are used To calculate the abnormality degree about amplitude p1.Figure 18 A is the chart for indicating the amplitude p1 calculated in this evaluation test.In Figure 18 A, Horizontal axis indicates that the number of wobbling action, the longitudinal axis indicate amplitude p1.As shown in Figure 18 A like that, in number of oscillations 1~57, amplitude p1 It is about 2~7, in contrast, after number of oscillations the 58th time, amplitude p1 increases to 8 or more.So, in number of oscillations After 58 times, influence brought by no vertical plate 52 can be seen in the value of amplitude p1.
Figure 18 B is the chart for indicating the abnormality degree about amplitude p1 calculated in this evaluation test.In Figure 18 B, horizontal axis is indicated The number of wobbling action, the longitudinal axis indicate abnormality degree.As shown in figure 18b, at the 1~57th time, on the basis of abnormality degree value hereinafter, In contrast, at the 58~64th time, abnormality degree is more than a reference value.Without vertical plate 52 after the 58th time, abnormality degree is correctly anti- The variation of the welded condition is reflected.
<evaluation test 7>
Inventors verify whether that burn-through can be detected by welded condition determination method involved in present embodiment.Workpiece, Hunting frequency and data number N and evaluation test 3 are same.
Figure 19 A is the chart for indicating the amplitude p1 calculated in this evaluation test.In fig. 19 a, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate amplitude p1.As shown in Figure 19 A like that, in number of oscillations 1~18, amplitude p1 is about 2~6, in contrast, In After number of oscillations the 19th time, influenced and amplitude p1 suddenly increasing by burning.
Figure 19 B is the chart for indicating the abnormality degree calculated in this evaluation test.In fig. 19b, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate abnormality degree.As shown in Figure 19 B like that, at the 1st~17th time, on the basis of abnormality degree value hereinafter, in contrast, At the 18th~19th time, abnormality degree is more than a reference value.In this way, by welded condition determination method involved in present embodiment, It knows that the variation for being detected as welded condition can will be burnt.
(embodiment 5)
In the present embodiment, welded condition decision-making system is obtaining each period swung in every swing period acquirement welding current The average value of value, the welding current value waved at end swung and the welding current value to be changed with swing period same period Amplitude this 3 characteristic quantities, abnormality degree is calculated to these characteristic quantities respectively, determines welded condition based on 3 abnormality degrees.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, due to illustrate same in embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 20 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~112 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
CPU111 calculates the average current in the i period as the 1st characteristic quantity (step S5131), extracts left end electric current from measured value And right end electric current is as the 2nd characteristic quantity (step S5132), make function to be repeated cyclically with swing period same period with The waveform of welding current value is adapted to, and the amplitude p1 of the function calculated is as the 3rd characteristic quantity (step S5133).Average current Calculating, the extraction of left end electric current and right end electric current and the calculating of amplitude p1, respectively with illustrate in embodiment 1 to 3 Equally.In addition, the processing of step S114~S115 due to illustrate same in embodiment 1, the description thereof will be omitted.
CPU111 calculates the abnormality degree about the 1st characteristic quantity as the 1st abnormality degree (step S5161), calculates about the 2nd characteristic quantity Abnormality degree as the 2nd abnormality degree (step S5162), calculate the abnormality degree about the 3rd characteristic quantity as the 3rd abnormality degree (step S5163).Illustrate in the calculating and embodiment 1 to 3 of 1st to the 3rd abnormality degree same.
Next, CPU111 is based respectively on the 1st to the 3rd abnormality degree to determine welded condition (step S517).In this process, 1st to the 3rd abnormality degree is compared with a reference value by CPU111 respectively, if 2 or more abnormality degrees are more than a reference value, welds shape State is exactly exception, is judged to needing to weld the stopping of movement.If the 1st to the 3rd abnormality degree all on the basis of value hereinafter, welded condition It is exactly normally, to be judged to not needing the stopping of welding movement.
The processing of step S118~S119 due to illustrate same in embodiment 1, the description thereof will be omitted.
By constituting as described above, welded condition decision-making system 100 by the 1st to the 3rd abnormality degree 2 or more be more than Welded condition is determined as exception in the case where a reference value, can be carried out the abnormal determination of high reliablity.
(embodiment 6)
In the welded condition determination method involved in above-mentioned embodiment 1 to 5, the exception that standard deviation is utilized is calculated Degree.Relevant abnormality degree cannot then be calculated in the characteristic quantity that normal distribution is not shown premised on characteristic quantity shows normal distribution Correct abnormality degree out.Such as the average current Iave in evaluation test 2 shows in number of oscillations 1~72 like that as shown in Figure 10 A Gentle increase tendency out.It shows in Figure 21 about the straight of number of oscillations the 1st~70th time average current in evaluation test 2 Fang Tu.In Figure 21, the longitudinal axis indicates data number (occurrence frequency), and horizontal axis indicates average current.In this example, in 272A and Normal distribution is not shown there are 2 peak values in data number in 273.5A.In addition, the standard deviation in the example is 0.9579. This is because the increase in average current in having time is inclined to.
Thus in the present embodiment, the time fluctuation linear approximation of characteristic quantity is calculated observation by welded condition decision-making system And the average and standard deviation of the difference of the difference of approximation, past value and approximation makes to be used to calculate abnormality degree.Thus special The difference of sign amount and approximation becomes close to normal distribution, can obtain the abnormality degree for correctly reflecting welded condition.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, same due to illustrating with embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 22 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~115 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
In welding movement, pass through the feedback control of robot controller 30 sometimes, observation gently changes, to target State is close.In order to cope with such gentle variation, the time fluctuation of characteristic quantity (average current) is carried out in present embodiment Linear approximation.
CPU111 is by the time fluctuation linear approximation (step S6161) of the average current in each period of wobbling action, based on close Abnormality degree (step S6162) is calculated like value.In this process, according to the difference of observation and approximation, past value and approximation The average and standard deviation of difference calculate abnormality degree.It is described in detail below.
Approximate expression Y=at+b is used to average current Iave (i), by the time fluctuation linear approximation of average current Iave (i).In This, Y is to carry out approximate average current, and t is the time, and a and b are parameters.The difference of observation and the approximation of average current exists Formula (21) provides.
[mathematical expression 11]
WhereinIt is the linear approximation of average current.
The average of the difference of the approximation of past value and average current provides in formula (22), and standard deviation is provided in formula (23).
[mathematical expression 12]
It is provided using its abnormality degree in formula (24).
[mathematical expression 13]
By in this way, the abnormality degree for excluding the influence of the gentle time fluctuation of the characteristic quantity under feedback control can be obtained.
The processing of step S117~S119 due to illustrate same in embodiment 1, the description thereof will be omitted.
By constituting as described above, can not be calculated just with being influenced by the gentle time fluctuation of the characteristic quantity under feedback control The really abnormality degree of reflection welded condition, can accurately determine welded condition.
So, the computer 101 of this structure is also as by the time fluctuation of characteristic quantity, approximately approximation portion is functioned, concurrently Wave function so that the approximation for the observation that calculation section above-mentioned is obtained based on observation and by above-mentioned approximate portion difference and mistake The average and standard deviation of the difference of the approximation for the past value for going value and above-mentioned approximate portion to obtain calculates abnormality degree.This is rear It is also the same in the embodiment 7,8 stated.
<evaluation test 8>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.In this evaluation test, abnormality degree is calculated using the data of average current identical with evaluation test 2.
Linear approximation is carried out to the time fluctuation of average current.Figure 23 is the chart for indicating the result of linear approximation.In Figure 23, Horizontal axis indicates that the number of wobbling action, the longitudinal axis indicate average current.As shown in figure 23 like that, in number of oscillations 1~70, average electricity Stream shows increase tendency in time, can carry out linear approximation to the time fluctuation.
The average and standard deviation for calculating the approximation of such linear approximation and the difference of past value, calculates different according to them Normal manner.Figure 24 A is the histogram of the difference of average current and approximation.In Figure 24 A, the longitudinal axis indicates data number (occurrence frequency), The difference of horizontal axis expression average current and approximation.In the histogram shown in Figure 24 A, more connect compared with the histogram shown in Figure 21 Nearly normal distribution.Standard deviation is reduced to 0.7433 from 0.9579, it is known that can lower deviation.
Figure 24 B is the chart for indicating the abnormality degree calculated in this evaluation test.In Figure 24 B, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate abnormality degree.Compared with not carrying out the abnormality degree of evaluation test 2 of linear approximation (with reference to Figure 10 B), in Figure 24 B Shown in result, the abnormality degree in number of oscillations 1~72 is reduced.Especially the 56th and the 64th time abnormality degree is lower, more Correctly reflect welded condition.In addition, abnormality degree is more than a reference value at the 73rd~76th time, control can correctly detect welding The variation of state.In addition, abnormality degree is lower than a reference value once again after the 77th time.It is thought that because, welding tack welding 54 The high data of partial abnormality degree are included in past value.
(embodiment 7)
Welded condition decision-making system 100 involved in present embodiment, is averaged in the same manner as embodiment 6 to as characteristic quantity The time fluctuation of electric current carries out linear approximation, and is more than the same as that of the second embodiment given a reference value by past abnormality degree Characteristic quantity is except past value, to calculate average and standard deviation, calculates abnormality degree using the average and standard deviation.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, same due to illustrating with embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 25 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~115 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
If average current when being more than a reference value comprising the abnormality degree calculated in the past in N number of past value, CPU111 are just removed (step S7161) outside.In the case where abnormality degree is more than a reference value, it is abnormal for capable of being estimated as welded condition.Therefore, by by this The abnormality degree of sample it is high data removal, in past value only comprising welded condition be normal time point under average current.
Next, CPU111 carries out linear approximation (step to the time fluctuation for the average current for eliminating the high data of abnormality degree S7162), calculated based on approximation abnormality degree (step S7163).In this process, according to the difference of observation and approximation, mistake The average and standard deviation of the difference of value and approximation is gone to calculate abnormality degree.It is described below.
It is set as the average current in the i-th-N to (i-1)-th average current Iave there are M abnormality degree more than a reference value.It will The data of this M average current be set to I ' ave (i, 1), I ' ave (i, 2), I ' ave (i, 3) ..., I ' ave (i, M), will Summarizing for they is set as I ' ave (i).By I ' ave (i) linear approximation.The difference of observation and the approximation of average current is in formula (25) it provides.
[mathematical expression 14]
WhereinIt is the linear approximation of average current.
The average of the difference of the approximation of past value and average current provides in formula (26), and standard deviation is provided in formula (27).
[mathematical expression 15]
It is provided using their abnormality degree in formula (28).
[mathematical expression 16]
In this way, the abnormality degree for excluding the influence of the gentle time fluctuation of the characteristic quantity under feedback control can be obtained.
It is same about illustrating in the processing of step S117~S118, with embodiment 1.In addition, in the present embodiment, with reality It is same to apply mode 2, after the notice occurred extremely, CPU111 does not send welding and processing is passed to step with stopping request S115.Also continue the calculating of abnormality degree and the judgement of welded condition in the case where being determined as that welded condition is exception as a result,.
By constituting as described above, characteristic quantity when due to being exception by welded condition, can not be by except past value The abnormality degree for influencing ground and calculating correct reflection welded condition of past abnormal characteristic quantity.In addition, passing through the exception to except The characteristic quantities of data carry out linear approximation, the difference of characteristic quantity and approximation becomes close to normal distribution, can correctly reflect weldering Connect the abnormality degree of state.
<evaluation test 9>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.In this evaluation test, abnormality degree is calculated using the data of average current identical with evaluation test 2.
Average current when by abnormality degree being more than a reference value except past value, to except the characteristic quantity of abnormal data when Between change carry out linear approximation.The average and standard deviation for calculating the difference of the obtained approximation and past value, according to it Calculate abnormality degree.Figure 26 is the chart for indicating the abnormality degree calculated in this evaluation test.In Figure 26, horizontal axis indicates to swing The number of movement, the longitudinal axis indicate abnormality degree.As shown in figure 26 like that, after number of oscillations the 73rd time, abnormality degree is more than benchmark Value, abnormality degree correctly reflect the variation of the welded condition under tack welding 54.In addition, in evaluation test 7, not about abnormality degree After the 77th time of a reference value, abnormality degree is nearly all more than a reference value.So, it is known that data that can not be high by abnormality degree The variation for influencing the correct detection welded condition in ground.
(embodiment 8)
The structure that linear approximation is carried out to the time fluctuation of characteristic quantity is described in above-mentioned embodiment 6.But in robot In the feedback control of control device 30, in the case that the difference of such target value and current observation is big after the rigid starting the arc, observation It is worth change dramatically.At this moment time fluctuation and non-rectilinear, cannot precisely carry out approximation in linear approximation.
For this purpose, in the present embodiment, welded condition decision-making system is not in the case where characteristic quantity changes linearly to its time Change and carry out that 2 ranks are approximate, come calculate the difference of observation and approximation, past value and approximation difference be averaged and standard deviation Difference calculates abnormality degree using them.
<structure of welding system>
The structure of welding system involved in present embodiment is as same with the structure of welding system 10 involved in embodiment 1 Sample, therefore same reference numerals are marked to identical constituent element, the description thereof will be omitted.
<movement of welding system>
The movement of welding system involved in present embodiment will be illustrated next.It is filled about robot controller 30 and power supply The movement for setting 40, due to illustrate same in embodiment 1, the description thereof will be omitted.
The movement of welded condition decision-making system 100 is illustrated.Figure 27 is to indicate welded condition involved in present embodiment The flow chart of the running order of decision-making system 100.The processing of step S111~115 due to illustrate in embodiment 1 it is same, Therefore the description thereof will be omitted.
CPU111 carries out 2 ranks approximation (step S8161) to the time fluctuation of the average current in each period of wobbling action, is based on Approximation calculates abnormality degree (step S8162).In this process, by the difference of observation and approximation, past value with it is approximate The average and standard deviation of the difference of value calculates abnormality degree.It is described in detail below.
Approximate expression Y=a (t is used to average current Iave (i)0-t)2+ b, it is close to 2 rank of time fluctuation of average current Iave (i) Seemingly.Here, Y is to carry out approximate average current, t0It is the time point determined, t is the time, and a and b are parameters.Observation The difference of value and the approximation of average current is provided in formula (29).
[mathematical expression 17]
WhereinIt is 2 rank approximations of average current.
The average of the difference of the approximation of past value and average current provides in formula (30), and standard deviation is provided in formula (31).
[mathematical expression 18]
It is provided using their abnormality degree in formula (32).
[mathematical expression 19]
In this way, the abnormality degree for excluding the influence of the time fluctuation of 2 functions of the characteristic quantity under feedback control can be obtained.
The processing of step S117~S119 due to illustrate same in embodiment 1, the description thereof will be omitted.
By constituting as described above, can not be calculated with being influenced by the nonlinear time fluctuation of the characteristic quantity under feedback control The abnormality degree of correct reflection welded condition, can accurately determine welded condition.
<evaluation test 10>
Inventors carry out the evaluation test of welded condition determination method involved in present embodiment as described below, verify it Performance.In this evaluation test, the workpiece for implementing tack welding in the position close to starting the arc place is welded, is verified whether Tack welding can be detected as to the variation of welded condition.In addition, hunting frequency and data number N are distinguished in this evaluation test It is set as 2Hz and 50 identically as evaluation test 2.In addition, being tested by gimmick same as evaluation test 1.
The chart of the time fluctuation of the average current after the rigid starting the arc is shown in Figure 28.In Figure 28, the longitudinal axis indicates average current, horizontal Axis indicates number of oscillations.In this example, at number of oscillations the 1st~30th time, average current is sharply reduced, which is It is nonlinear.In addition, tack welding is located near the 36th time, the influence of the rising characterization tack welding of the average current in this part.
It is in this evaluation test that 2 rank of the time fluctuation of average current is approximate.Figure 29 A is the chart for indicating the approximate result of 2 ranks. In Figure 29 A, horizontal axis indicates that the number of wobbling action, the longitudinal axis indicate average current.In addition, indicated by the solid line, to carry out 2 ranks close As average current time fluctuation.
The average and standard deviation for calculating the difference of such approximate approximation of 2 ranks and past value, calculates different according to them Normal manner.Figure 29 B is the chart for indicating the abnormality degree calculated in this evaluation test.In Figure 29 B, horizontal axis indicates time of wobbling action Number, the longitudinal axis indicate abnormality degree.It is more than benchmark in the 36th abnormality degree in the result shown in Figure 29 B.So, pass through we Tack welding can be detected as the variation of welded condition in electric arc from the starting the arc between stable state transfer by method.
(other embodiments)
It describes to act welding in the case where being determined as that welded condition is exception in above-mentioned embodiment 1 to 8 and stop Structure, but not limited to this.It can be set as the structure of the change welding control in the case where being determined as that welded condition is exception. Figure 30 is the explanatory diagram for illustrating the welding in the copying control to the workpiece for implementing tack welding.As shown in Figure 30 In the case of, by copying control, traveling is welded along the straight line of the cross part of sealing wire i.e. 2 component.If electric arc arrival is set Tack welding 54 in the midway of the cross part can accidentally detect on the boundary of tack welding 54 for sealing wire, such as then in copying control It is welded like that along the verge advance shown in arrow in figure.In order to eliminate such failure welding, moreover it is possible to be detected by tack welding In the case where exception for welded condition, continues the control of feedback control and wobbling action and stop copying control.
The approximate structure of 2 ranks is carried out to the time fluctuation in addition, describing in above-mentioned embodiment 8, but is not limited to This.It can also be set as not carrying out 2 ranks approximation to the time fluctuation of characteristic quantity, carry out the structure of exponential approximation.
In above-mentioned embodiment 6, the structure that linear approximation is carried out to the time fluctuation of characteristic quantity is described, in embodiment It is described in 8 and the approximate structure of 2 ranks is carried out to the time fluctuation.It can also be set as executing such linear approximation and 2 ranks are approximate The structure of both sides.In this case, since the arc welding after the rigid starting the arc is from unstable state to being passed to stable state Until, characteristic quantity sharply time fluctuation, therefore execution 2 ranks approximation executes linear close under the gentle stable state of time fluctuation Like.
In addition, describing the structure for using welding current value as physical quantity in above-mentioned embodiment 1 to 8, but not It is defined in this.As long as corresponding to the physical quantity of the cyclical swing of wobbling action, welding current value may not be.In The such welding using constant-current power supply of TIG (Tungsten Inert Gas, non-melt pole inert gas-shielded arc welding) welding In method, the periodic weld voltage value corresponding to wobbling action changes.For this purpose, welding voltage value can be used as physical quantity. In addition, the structure of average current, left end electric current and right end electric current and the amplitude of welding current value as characteristic quantity is described, But not limited to this.Average value in the swing period of welding current value, the welding gun in wobbling action can also be waved end The welding current value at place, using the amplitude of the welding current value changed with swing period same period as characteristic quantity.In addition, can also By the integral of the physical quantity in the physical quantity for waving centre of the welding gun in wobbling action or the fixed interval comprising waving end Value is used as characteristic quantity.The integrated value is illustrated.Figure 31 A and Figure 31 B are for illustrating the fixed interval comprising waving end In welding current integrated value chart.In Figure 31 A and Figure 31 B, horizontal axis indicates the time, and the longitudinal axis indicates welding current Value.In addition, label is that the peak value of R " is corresponding with right end in figure, it is corresponding with left end labeled as the peak value of " L ".In Figure 31 A, wave The part of the paddy of shape is central corresponding with waving in wobbling action.Thus, between adjacent 2 paddy of the peak value comprising " R " Section is to return to (hereinafter referred to as " the right side of the section until waving center once again to passing through right end from the center of waving of wobbling action Between lateral areas "), the section between adjacent 2 paddy of the peak value comprising " L " is returned once again to by left end from waving center Wave the section (hereinafter referred to as " left side section ") until center.The integrated value of right side section or the welding current in the section of left side SR, SL (oblique line portion) are changed due to corresponding to swing period, are the values for reflecting welded condition, can be made as characteristic quantity With.Alternatively, it is also possible to not using the initial point in section and terminal as waving center.Such as it can also determine as shown in figure 31b Welding current value is become the integrated value in the section of a reference value or more as characteristic quantity by a reference value of welding current value.
In addition, describing in above-mentioned embodiment 5 such as flowering structure: obtaining weldering in each swing period in each period of swing The average value for connecing current value, the welding current value waved at end swung and the welding to be changed with swing period same period These characteristic quantities are calculated abnormality degree by this 3 characteristic quantities of the amplitude of current value respectively, determine to weld shape based on 3 abnormality degrees State, but not limited to this.It can be set as such as flowering structure: obtain average value in the swing period of physical quantity, in wobbling action The physical quantity of welding gun waved at end, the physical quantity for waving center of welding gun in wobbling action, comprising wave end section in The integrated value of physical quantity or the amplitude of the physical quantity to be changed with swing period same period at least two it is above as Characteristic quantity calculates abnormality degree to these characteristic quantities respectively, determines welded condition based on multiple abnormality degrees.
In addition, describing in above-mentioned embodiment 1 to 8 and executing welded condition decision procedure by single computer 101 150 structure all handled, but the present invention is not limited to this, moreover it is possible to be set as being dispersed to execute by multiple devices (computer) with The decentralized system that welded condition decision procedure 150 is similarly handled.In addition, describing individual setting welderings in embodiment 1 to 8 The structure of state decision-making system 100 and robot controller 30 is connect, but not limited to this.It can also determine in welded condition Control, copying control and the welding that the function that system 100 is provided as robot controller 30 executes wobbling action are dynamic The function of the feedback control of work and the function of controlling power supply device 40.
Industrial availability
Welded condition decision-making system and welded condition determination method of the invention is as the welded condition determined in arc welding Welded condition decision-making system and welded condition determination method etc. and it is useful.
This application is based on Japanese patent application (Patent 2017-47262) filed on March 13rd, 2017, by its content at this In be hereby incorporated by reference into.
The explanation of appended drawing reference
10 welding systems
20 welding robots
21 welding guns
23 wire feed units
24 welding wires
30 robot controllers
301 CPU
330 control programs
40 power supply devices
41 constant voltage power supplies
42 welding current test sections
43 switches
50 workpiece
100 welded condition decision-making systems
111 CPU
114 hard disks
130 display units
150 welded condition decision procedures

Claims (13)

1. a kind of welded condition decision-making system, to the welding shape in the arc welding under the wobbling action for making welding gun periodical wobble State is determined that the welded condition decision-making system is characterized in that having:
Acquisition unit obtains relevant to the physical quantity for corresponding to the period and changing in each period of the wobbling action Characteristic quantity；
Calculation section calculates abnormality degree, the sight based on the average and standard deviation about observation and multiple past values Measured value is the characteristic quantity obtained by the acquisition unit a period, and the past value is by the acquisition unit in institute State the characteristic quantity obtained before a cycle；With
Determination unit determines the welded condition based on the abnormality degree calculated by the calculation section.
2. welded condition decision-making system according to claim 1, which is characterized in that
The calculation section is configured to calculate the abnormality degree in each period,
The welded condition decision-making system is also equipped with:
Except outside, the abnormality degree calculated in the period before one period the situation higher than given a reference value Under, will be described before period in the abnormality degree calculating in be used as the characteristic quantity of the observation from the multiple Except past value.
3. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The welded condition decision-making system is also equipped with:
Approximate portion, the time fluctuation of the characteristic quantity is approximate,
The calculation section be configured to the approximation of the observation obtained based on the observation and by the approximate portion difference, It is calculated with the average and standard deviation of the difference of the approximation of the past value and the past value obtained by the approximate portion The abnormality degree out.
4. welded condition decision-making system according to claim 3, which is characterized in that
The approximation portion is configured to by linear approximation that the time fluctuation of the characteristic quantity is approximate.
5. welded condition decision-making system according to claim 3, which is characterized in that
The approximation portion is configured to approximate by 2 ranks or exponential approximation is approximate by the time fluctuation of the characteristic quantity.
6. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The characteristic quantity is that average value in the period of the physical quantity, the welding gun in the wobbling action wave The physical quantity for waving centre of the welding gun in the physical quantity, the wobbling action at end is held comprising waving Section in the physical quantity integrated value or amplitude with the approximate waveform of periodical time fluctuation of the physical quantity.
7. welded condition decision-making system according to claim 3, which is characterized in that
The characteristic quantity is that average value in the period of the physical quantity, the welding gun in the wobbling action wave The physical quantity for waving centre of the welding gun in the physical quantity, the wobbling action at end is held comprising waving Section in the physical quantity integrated value or amplitude with the approximate waveform of periodical time fluctuation of the physical quantity.
8. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The acquisition unit is constituted are as follows: obtains average value in the period of the physical quantity, described in the wobbling action The physical quantity of welding gun waved at end, the physical quantity for waving centre of the welding gun in the wobbling action, The integrated value of the physical quantity in section comprising waving end is approximate with the periodical time fluctuation of the physical quantity In the amplitude of waveform more than at least two, using as the characteristic quantity,
The calculation section is configured to calculate the abnormality degree respectively to each characteristic quantity,
The determination unit is configured to determine the welded condition based on each abnormality degree calculated by the calculation section.
9. welded condition decision-making system according to claim 3, which is characterized in that
The acquisition unit is constituted are as follows: obtains average value in the period of the physical quantity, described in the wobbling action The physical quantity of welding gun waved at end, the physical quantity for waving centre of the welding gun in the wobbling action, The integrated value of the physical quantity in section comprising waving end is approximate with the periodical time fluctuation of the physical quantity In the amplitude of waveform more than at least two, as the characteristic quantity,
The calculation section is configured to calculate the abnormality degree respectively to each characteristic quantity,
The determination unit is configured to determine the welded condition based on each abnormality degree calculated by the calculation section.
10. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The welded condition decision-making system is also equipped with:
Control unit controls the movement of the arc welding based on the judgement result of the welded condition of the determination unit.
11. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The welded condition decision-making system is also equipped with:
Exceptional communication portion notifies the exception in the case where being determined as that the welded condition is exception by the determination unit.
12. welded condition decision-making system according to claim 1 or 2, which is characterized in that
The physical quantity is welding current or weldingvoltage.
13. a kind of welded condition determination method, to the welding in the arc welding under the wobbling action for making welding gun periodical wobble State is determined that the welded condition determination method is characterized in that, has following steps:
Characteristic quantity relevant to the physical quantity changed corresponding to the period is obtained in each period of the wobbling action；
Abnormality degree is calculated based on the average and standard deviation about observation and multiple past values, the observation is one The characteristic quantity obtained in a period, the past value are the characteristic quantities obtained before one period； With
The welded condition is determined based on the abnormality degree of calculating.
CN201880015443.4A 2017-03-13 2018-02-27 Welding state determination system and welding state determination method Active CN110446577B (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP2017047262A JP6735697B2 (en) 2017-03-13 2017-03-13 Welding state determination system and welding state determination method
JP2017-047262 2017-03-13
PCT/JP2018/007369 WO2018168448A1 (en) 2017-03-13 2018-02-27 Welding state determination system and welding state determination method

Publications (2)

Publication Number Publication Date
CN110446577A true CN110446577A (en) 2019-11-12
CN110446577B CN110446577B (en) 2021-04-20

Family

ID=63522376

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201880015443.4A Active CN110446577B (en) 2017-03-13 2018-02-27 Welding state determination system and welding state determination method

Country Status (6)

Country Link
US (1) US20200009673A1 (en)
EP (1) EP3597345A4 (en)
JP (1) JP6735697B2 (en)
KR (1) KR102202535B1 (en)
CN (1) CN110446577B (en)
WO (1) WO2018168448A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114178653A (en) * 2021-12-23 2022-03-15 唐山松下产业机器有限公司 Welding quality detection method and system

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11897060B2 (en) * 2017-11-29 2024-02-13 Lincoln Global, Inc. Systems and methods for welding torch weaving
US11065707B2 (en) * 2017-11-29 2021-07-20 Lincoln Global, Inc. Systems and methods supporting predictive and preventative maintenance
JP7251988B2 (en) * 2019-01-22 2023-04-04 株式会社神戸製鋼所 PROFILE CONTROL METHOD FOR PULSE ARC WELDING, CONTROL DEVICE, WELDING SYSTEM, WELDING PROGRAM AND WELDING POWER SUPPLY
KR102306269B1 (en) * 2020-04-24 2021-09-30 부산대학교 산학협력단 Method and apparatus for real-time arc-welding defect detection and classification

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0112579A1 (en) * 1982-12-27 1984-07-04 Hitachi, Ltd. Method and apparatus for welding line tracer control
EP0186041A1 (en) * 1984-12-21 1986-07-02 Siemens Aktiengesellschaft Process and appliance for guiding arc-welding guns of mechanised welding plants
US6118191A (en) * 1996-12-19 2000-09-12 Denso Corporation Method for bonding rotors of an electric rotating machine
CN101422837A (en) * 2007-10-31 2009-05-06 株式会社神户制钢所 Arc welding robot control system and method thereof
CN102151949A (en) * 2010-12-29 2011-08-17 郑州华创机电技术有限公司 Method and device for bidirectional automatic tracking by argon arc welding arc oscillation of tungsten electrode
CN102615390A (en) * 2012-04-28 2012-08-01 湘潭大学 Swing arc-based multi-layer and multi-channel weld tracking system and identification method thereof
CN103182589A (en) * 2011-12-30 2013-07-03 中国科学院沈阳自动化研究所 Arc oscillation-based gas shielded welding arc tracking method
CN105665885A (en) * 2016-04-07 2016-06-15 湘潭大学 Vertical weld tracking system and method capable of adaptively adjusting inclination angle of welding gun

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2570015B2 (en) * 1991-08-28 1997-01-08 日本鋼管株式会社 Method and apparatus for determining welding stability
JP3906561B2 (en) 1998-04-27 2007-04-18 石川島播磨重工業株式会社 Welding quality judgment method and apparatus
KR20030049325A (en) * 2001-12-14 2003-06-25 삼성전자주식회사 Apparatus and contrlomethod of arcwelding system
JP4728852B2 (en) * 2006-03-20 2011-07-20 エスペック株式会社 Bonding evaluation method and apparatus
JP4891726B2 (en) * 2006-10-06 2012-03-07 株式会社神戸製鋼所 Robot controller for controlling tandem arc welding system and arc scanning control method using the same
WO2010135369A1 (en) 2009-05-18 2010-11-25 Dose Medical Corporation Drug eluting ocular implant
US8569646B2 (en) * 2009-11-13 2013-10-29 Lincoln Global, Inc. Systems, methods, and apparatuses for monitoring weld quality
JP6052798B2 (en) * 2013-07-08 2016-12-27 株式会社神戸製鋼所 Abnormality monitoring device for automatic welding machine
KR101779736B1 (en) * 2014-06-12 2017-09-18 닛산 지도우샤 가부시키가이샤 Bonding state inspection method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0112579A1 (en) * 1982-12-27 1984-07-04 Hitachi, Ltd. Method and apparatus for welding line tracer control
EP0186041A1 (en) * 1984-12-21 1986-07-02 Siemens Aktiengesellschaft Process and appliance for guiding arc-welding guns of mechanised welding plants
US6118191A (en) * 1996-12-19 2000-09-12 Denso Corporation Method for bonding rotors of an electric rotating machine
CN101422837A (en) * 2007-10-31 2009-05-06 株式会社神户制钢所 Arc welding robot control system and method thereof
CN102151949A (en) * 2010-12-29 2011-08-17 郑州华创机电技术有限公司 Method and device for bidirectional automatic tracking by argon arc welding arc oscillation of tungsten electrode
CN103182589A (en) * 2011-12-30 2013-07-03 中国科学院沈阳自动化研究所 Arc oscillation-based gas shielded welding arc tracking method
CN102615390A (en) * 2012-04-28 2012-08-01 湘潭大学 Swing arc-based multi-layer and multi-channel weld tracking system and identification method thereof
CN105665885A (en) * 2016-04-07 2016-06-15 湘潭大学 Vertical weld tracking system and method capable of adaptively adjusting inclination angle of welding gun

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114178653A (en) * 2021-12-23 2022-03-15 唐山松下产业机器有限公司 Welding quality detection method and system

Also Published As

Publication number Publication date
KR102202535B1 (en) 2021-01-13
WO2018168448A1 (en) 2018-09-20
EP3597345A1 (en) 2020-01-22
JP2018149567A (en) 2018-09-27
CN110446577B (en) 2021-04-20
JP6735697B2 (en) 2020-08-05
EP3597345A4 (en) 2021-03-24
US20200009673A1 (en) 2020-01-09
KR20190117592A (en) 2019-10-16

Similar Documents

Publication Publication Date Title
CN110446577A (en) Welded condition decision-making system and welded condition determination method
CN106392266B (en) Rote learning device, electric arc welding controller and arc welding robot system
CN106457468B (en) Use the hybrid laser welding system and method for Liang Ge robot
US20150273604A1 (en) Material joining inspection and repair
CN103358308B (en) Robotic gun orientation standardization
CN106994545B (en) Setting support device and setting support method
CN105636734A (en) Extraction of arc length from voltage and current feedback
CN108883495A (en) The method and device of seam is especially monitored when being engaged by means of laser emission
US20030111450A1 (en) Arc welding apparatus and control method thereof
JP6794685B2 (en) Image processing method and image processing equipment
WO2020251038A1 (en) Repair weld control device and repair weld control method
JP2015225214A (en) Simulation system
JP2020203308A (en) Control device, program, and robot control system
KR20160020698A (en) Method for quality judgment to spot welding
WO2023234291A1 (en) Welding system and welding method
CN105537820A (en) Welding system and welding method
ES2969093T3 (en) Welding control device, welding control procedure and welding control program
KR100762365B1 (en) Seam tracking method of welding robot based on probability model considered the property of &#39;through-arc&#39; sensor
US10232457B2 (en) Arc welding robot system
KR101584421B1 (en) Monitoring system for arc welding
JP3146943B2 (en) Arc welding robot system
WO2022092061A1 (en) Robotic welding system
US20240112596A1 (en) Weld training systems with welding technique monitoring
Pires et al. Sensors for Welding Robots
JPS63224873A (en) Sensing parameter automatic measuring and storing system for arc sensor

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant